This study investigated the association between migraine severity and sleep quality in all regions of Saudi Arabia from August to October 2023. This study is a comparative cross-sectional study enrolling 1,399 participants from all regions of Saudi Arabia. The required sample size was calculated using the Epitools website¹ using g-power, with an effect size of 0.1 and one degree of freedom. The inclusion criterion was anyone above the age of 18 years of either sex. The exclusion criteria included primary headaches other than migraine, evaluation using the IHS criteria, and surgery within the last 6 months. The participants were further divided into migraine patients (cases) and non-migraine patients (controls), based on the application of the IHS criteria.

Data were obtained using an online, structured, self-administered questionnaire in Arabic created using Google Forms. The first part of the questionnaire explored the sociodemographic characteristics of the study participants. Age, sex, height, weight, and BMI of the participants were all obtained. Other participant characteristics, such as smoking status, presence of comorbidities, and family history of headaches, were also recorded. Finally, a screening question regarding the history of headaches in one’s life was posed to all participants. If the participant answered yes, they were moved to the IHS criteria page and the Migraine Disability Assessment Scale (MIDAS). If not, they would move on to the Pittsburgh Sleep Quality Index (PSQI) page.

The International Headache Society (IHS) criteria were included in the questionnaire to diagnose and differentiate patients with migraine from non-migraine patients (control group). The IHS classification is a 4-digit hierarchical method used to diagnose headaches with variable specificities, with the first digit being specific for diagnostic categories, the second digit for subtypes, and the third and fourth digits for more specifications. The criteria were as follows: untreated or unsuccessfully treated headache attacks that last 4–72 h occurring more than 15 times per month not attributed to any other disease; a headache that had at least two features (unilateral, pulsatile, moderate, and severe pain, aggravated by physical activity); and at least one of the following associated symptoms (nausea and/or vomiting, photophobia, or phonophobia) must occur alongside the headache (30).

An Arabic-validated form of the MIDAS questionnaire was used to measure disability levels linked to migraine (31). Comprising five questions, this questionnaire assesses the number of workdays lost due to migraine over 3 months. The MIDAS focuses on migraine-induced disability affecting work/study, household duties, and leisure activities on migraine days (32). The questions measure either complete inactivity days or days with productivity reduced by 50% or more. The cumulative total of these days was subsequently categorized into four disability severity levels, as follows: Grade I (0–5 points), indicating little to no disability; Grade II (6–10 points), mild disability; Grade III (11–20 points), moderate disability; and Grade IV (21 points or more), severe disability. The MIDAS questionnaire has two components: MIDAS A, which tracks the frequency of headaches, and MIDAS B, which rates the intensity of pain on a scale of 0 (no pain) to 10 (extremely severe pain) over a three-month period.

The PSQI is a questionnaire that assesses total sleep time, self-reported sleep quality, and disorders present over the past month. The questionnaire contains 19 questions distributed across seven domains. Subjective sleep quality, sleep latency, sleep length, sleep activity, sleep medication usage, sleep disorders, and daytime functional performance are the seven primary elements of this scale. The sum of all component scores yields the total PSQI score, for which a decline in sleep quality is indicated by a PSQI score of >5. The Arabic version of the questionnaire was based on a previously published paper (33) and incorporated into Google Forms. Respondents were either asked to write a short text answer or select one out of four options, as per the PSQI questionnaire.

Statistical Analysis was conducted using the Statistical Package for the Social Sciences (SPSS) version 29.0, released in 2022 by SPSS Inc., Chicago, IL, United States. Data are presented as the mean ± standard deviation (SD), or frequency counts and corresponding percentages. Chi-square and independent t-tests were applied to compare variables. Pearson’s correlation analysis was further used to assess the correlations between the scales. A multivariate linear regression analysis was conducted to evaluate the effect of sleep quality on migraine severity under different adjustment conditions and subgroups. Likewise, multivariate linear regression analysis was conducted to evaluate the effect of migraine severity on sleep quality. Three multivariate linear regression models were used, and all variables that were significant in the univariate regression analysis were included in the multivariate regression models. Statistical significance was set at p < 0.05.

A total of 1,567 individuals completed the questionnaire, of whom 35 were excluded for being under 18 years of age, 116 for having undergone a surgical procedure within the last 6 months, and 17 for disagreeing to participate in the study. Finally, the enrolled cohort comprised 1,399 individuals with a mean (SD) age of 27.9 ± 11.6 and BMI of 24.4 ± 6.1, most of whom were from the central region of Saudi Arabia (376 [26.9%]), were married (977 [69.8%]), and were university students or graduates (1,056 [75.5%]). A small proportion of participants were current or previous smokers (n = 129 [9.1%]). Moreover, almost two-thirds of the study participants had a family member complaining of headaches (898 [71.7%]).

Of the 1,399 participants included in the study, 349 had migraine according to the IHS diagnostic criteria, representing 24.95% of the total study population. Table 1 presents the sociodemographic data of the migraine patients and 1,050 controls. Notable findings included a significant sex difference, with females predominating in the migraine group [311 (89.1) vs. 758 (72.2); p < 0.001]. A higher proportion of individuals in the migraine group were unemployed or retired than in the control group [103 (29.5) vs. 208 (19.8); p < 0.001]. The average height of the migraine group was notably lower than that of the control group [160.14 (7.9) vs. 162.6 (9.2); p < 0.001].

As shown in Table 2, of the 349 participants with migraine, 83% did not miss work or school because of headaches over a period of 3 months. However, more than half of the participants did not perform household work for one or more days because of headaches. The mean total MIDAS score was 8.19 days, and was shown to be negatively affected by headaches for a duration of 3 months, while the mean MIDAS B score was 7.05. Regarding the overall MIDAS grade, more than half of the participants belonged to the little or no disability group. Moreover, only approximately 10% of migraine patients were categorized in the severe disability group.

As shown in Table 3, there was no significant difference in PSQI scores between the migraine and control groups. The overall cohort had a score of 6.9 (3.4), with 7.02 (3.4) for the migraine group and 6.97 (3.4) for the control group. The subjective sleep quality scores were 1.15 (0.8) in the migraine group and 1.09 (0.830) in the control group. The daily disturbance factor scores were similar between the migraine and control groups, with scores of 1.32 (0.64) and 1.36 (0.7), respectively. Overall, the PSQI score showed that most of the 813 participants had mild sleeping difficulty (56.4% migraine patients and 58.7% controls), and approximately 539 participants had moderate sleeping difficulty (40.1% migraine patients and 38% controls).

The association between MIDAS grade and various characteristics in the 349 migraine patients is presented in Table 4, of whom 259 had a little/mild MIDAS grade, and the remaining 90 migraine patients had moderate/severe disability levels. A family history of headache was significantly more common in both the little/mild (69.9%) and moderate/severe (76.7%) migraine severity groups. Another statistically significant parameter, shown in Table 4, was waking up from sleep due to headaches. More than half of those with a little/mild severity level (61.4%) did not have headache-related awakenings, whereas (35.5%) of those with the same severity grade reported sleep interruption due to headache. Conversely, approximately 50 percent of moderate/severe migraine patients (62.2%) reported waking from sleep, compared to 33.3 and 4.4% of the remaining patients who never wake up and always wake up because of headaches by (33.3%) and (4.4%), respectively. Among the medications taken for headache, ibuprofen was the only significant drug taken for headaches among both little/mild migraine patients (7.3%) and moderate/severe migraine patients (17.8%), as opposed to Panadol, Solpadeine, or other medications. Regarding whether headache improved with medications, both the little/mild (44.8%) and moderate/severe (65.6%) MIDAS groups reported that their headache sometimes improved with medications. Of the reported PSQI scores, the PSQI sleep latency score (C2; p = 0.02), PSQI sleep disturbance score (C5; p < 0.001), and PSQI daytime dysfunction score (C7; p = 0.022) all showed significant associations, as shown in Table 4. Finally, there was no statistically significant difference in PSQI scores across patients with different MIDAS severities.

Table 5 presents the association between the PSQI and migraine characteristics in 349 patients with migraine. Accordingly, 201 patients with migraine had no/mild PSQI grades and 148 had moderate/severe levels of sleep difficulty. Statistical analysis of the sociodemographic variables revealed no statistical differences in PSQI grade. High blood pressure was the only statistically significant comorbidity among patients with migraine. Statistically significant differences were further observed between those with no or mild sleeping difficulties, and those with moderate or severe sleeping difficulties taking medications for the comorbidities listed in Table 5. Among migraine patients in both PSQI groups, Table 5 shows the factors that trigger/worsen headaches. Overall, a lack of or excessive sleep was the most common migraine trigger for people with no/mild sleeping difficulties (61.7%) and moderate/severe sleeping difficulties (81.8%). This was followed by fatigue, stress, and certain smells with comparative frequencies in both the no/mild and moderate/severe PSQI groups, as shown in Table 5. Both PSQI groups reported significant results, indicating that both suffered from sleep disturbances as a result of headaches, albeit with differences in frequency. The no/mild sleeping difficulty group reported almost never waking up from sleep because of headaches. However, the moderate/severe group occasionally awoke from sleep. On the PSQI, 33.8% of those with no or mild sleeping difficulties reported occasional disturbances during sleep, compared with 54.1% of those with moderate or severe sleeping difficulties. In the no/mild sleep difficulty group, 3% reported always experiencing disturbed sleep due to headaches, whereas the moderate/severe sleep difficulty group reported a PSQI score of 4.1%. Medications improving headaches have been shown to be statistically significant in both PSQI groups, with the moderate to severe group reporting that using medications sometimes improved the headache (60.8%) With respect to the means to avoid headaches, taking medications demonstrated a statistical significance in both the no/mild sleeping difficulty (36.8%) and the moderate/ severe sleeping difficulty (50%) groups, making it the only statistically significant prophylactic strategy employed compared to the other practices. In this study, PSQI severity exhibited a significant p-value of <0.001. MIDAS scores and grades had p-values of 0.002 and 0.003, respectively, indicating that MIDAS grades and sleep quality were related.

A positive correlation was found between the PSQI and MIDAS scores in migraine patients, which was statistically significant (r = 0.18, p < 0.001) as shown in Table 6. The MIDAS was significantly correlated positively with the following PSQI components: subjective sleep quality (r = 0.11, p = 0.047), sleep latency (r = 0.19, p < 0.001), sleep disturbance (r = 0.23, p < 0.001), and daytime dysfunction (r = 0.2, p < 0.001). The Pearson correlation test did not reveal any association between the PSQI and its components with MIDAS B (Figure 1).

Multivariate linear regression analysis of sleep quality on migraine severity, represented as the MIDAS, is shown in Table 7. In total, three models were constructed based on the results of the univariate regression analysis, literature, and the clinical significance of these factors. This analysis revealed that the PSQI score under no-adjustment conditions, as in Model 1, was insignificant in predicting migraine severity. After adjusting for smoking, as in Model 2, a significant association between smoking and migraine severity was detected, whereby an incremental increase in smoking worsened migraine severity by 0.108 (beta coefficient = −0.108, p < 0.042). Model 3 further illustrated the significance of smoking as a predictor of migraine severity, despite adjusting for age, sex, and BMI.

Table 8 shows the results of the multivariate linear regression analysis. The MIDAS score was the only significant factor retained after univariate regression analysis for all demographics. Hence, all variables were adjusted in Model 2. A significant association between migraine severity and poor sleep quality in patients with migraine was detected in Models 1–2. For every 1-point increase in the MIDAS score (i.e., 1 day missed due to migraine headache), the PSQI score (i.e., sleeping difficulty) increased by 0.179 (beta coefficient = 0.179, p < 0.001) in all migraine patients.